blob: 077575fe1197f10a8dfee338cf49255141a9fe1c [file] [log] [blame]
Chris Lattnerc38ef542007-10-22 04:32:37 +00001<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3
4<html>
5<head>
Chris Lattner95ce0d62007-11-06 05:02:48 +00006 <title>Kaleidoscope: Tutorial Introduction and the Lexer</title>
Chris Lattnerc38ef542007-10-22 04:32:37 +00007 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
8 <meta name="author" content="Chris Lattner">
9 <link rel="stylesheet" href="../llvm.css" type="text/css">
10</head>
11
12<body>
13
Chris Lattner95ce0d62007-11-06 05:02:48 +000014<div class="doc_title">Kaleidoscope: Tutorial Introduction and the Lexer</div>
Chris Lattnerc38ef542007-10-22 04:32:37 +000015
Chris Lattner128eb862007-11-05 19:06:59 +000016<ul>
Chris Lattner0e555b12007-11-05 20:04:56 +000017<li><a href="index.html">Up to Tutorial Index</a></li>
Chris Lattner128eb862007-11-05 19:06:59 +000018<li>Chapter 1
19 <ol>
20 <li><a href="#intro">Tutorial Introduction</a></li>
21 <li><a href="#language">The Basic Language</a></li>
22 <li><a href="#lexer">The Lexer</a></li>
23 </ol>
24</li>
Chris Lattner0e555b12007-11-05 20:04:56 +000025<li><a href="LangImpl2.html">Chapter 2</a>: Implementing a Parser and AST</li>
Chris Lattner128eb862007-11-05 19:06:59 +000026</ul>
27
Chris Lattnerc38ef542007-10-22 04:32:37 +000028<div class="doc_author">
29 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
30</div>
31
32<!-- *********************************************************************** -->
33<div class="doc_section"><a name="intro">Tutorial Introduction</a></div>
34<!-- *********************************************************************** -->
35
36<div class="doc_text">
37
38<p>Welcome to the "Implementing a language with LLVM" tutorial. This tutorial
Chris Lattner95ce0d62007-11-06 05:02:48 +000039runs through the implementation of a simple language, showing how fun and
Chris Lattner619bc0a2007-11-05 20:13:56 +000040easy it can be. This tutorial will get you up and started as well as help to
Chris Lattner4134c282007-11-06 06:32:18 +000041build a framework you can extend to other languages. The code in this tutorial
42can also be used as a playground to hack on other LLVM specific things.
Chris Lattnerc38ef542007-10-22 04:32:37 +000043</p>
44
Chris Lattner95ce0d62007-11-06 05:02:48 +000045<p>
46The goal of this tutorial is to progressively unveil our language, describing
47how it is built up over time. This will let us cover a fairly broad range of
48language design and LLVM-specific usage issues, showing and explaining the code
49for it all along the way, without overwhelming you with tons of details up
50front.</p>
51
52<p>It is useful to point out ahead of time that this tutorial is really about
53teaching compiler techniques and LLVM specifically, <em>not</em> about teaching
54modern and sane software engineering principles. In practice, this means that
55we'll take a number of shortcuts to simplify the exposition. For example, the
56code leaks memory, uses global variables all over the place, doesn't use nice
57design patterns like visitors, etc... but it is very simple. If you dig in and
58use the code as a basis for future projects, fixing these deficiencies shouldn't
59be hard.</p>
60
61<p>I've tried to put this tutorial together in a way that makes chapters easy to
Chris Lattner4134c282007-11-06 06:32:18 +000062skip over if you are already familiar with or are uninterested in the various
63pieces. The structure of the tutorial is:
Chris Lattner95ce0d62007-11-06 05:02:48 +000064</p>
65
66<ul>
67<li><b><a href="#language">Chapter #1</a>: Introduction to the Kaleidoscope
68language, and the definition of its Lexer</b> - This shows where we are going
69and the basic functionality that we want it to do. In order to make this
70tutorial maximally understandable and hackable, we choose to implement
71everything in C++ instead of using lexer and parser generators. LLVM obviously
72works just fine with such tools, feel free to use one if you prefer.</li>
73<li><b><a href="LangImpl2.html">Chapter #2</a>: Implementing a Parser and
74AST</b> - With the lexer in place, we can talk about parsing techniques and
75basic AST construction. This tutorial describes recursive descent parsing and
76operator precedence parsing. Nothing in Chapters 1 or 2 is LLVM-specific,
77the code doesn't even link in LLVM at this point. :)</li>
78<li><b><a href="LangImpl3.html">Chapter #3</a>: Code generation to LLVM IR</b> -
79With the AST ready, we can show off how easy generation of LLVM IR really
80is.</li>
81<li><b><a href="LangImpl4.html">Chapter #4</a>: Adding JIT and Optimizer
82Support</b> - Because a lot of people are interested in using LLVM as a JIT,
83we'll dive right into it and show you the 3 lines it takes to add JIT support.
84LLVM is also useful in many other ways, but this is one simple and "sexy" way
Chris Lattner4134c282007-11-06 06:32:18 +000085to shows off its power. :)</li>
Chris Lattner95ce0d62007-11-06 05:02:48 +000086<li><b><a href="LangImpl5.html">Chapter #5</a>: Extending the Language: Control
87Flow</b> - With the language up and running, we show how to extend it with
Chris Lattner4134c282007-11-06 06:32:18 +000088control flow operations (if/then/else and a 'for' loop). This gives us a chance
Chris Lattner95ce0d62007-11-06 05:02:48 +000089to talk about simple SSA construction and control flow.</li>
90<li><b><a href="LangImpl6.html">Chapter #6</a>: Extending the Language:
91User-defined Operators</b> - This is a silly but fun chapter that talks about
92extending the language to let the user program define their own arbitrary
93unary and binary operators (with assignable precedence!). This lets us build a
94significant piece of the "language" as library routines.</li>
95<li><b><a href="LangImpl7.html">Chapter #7</a>: Extending the Language: Mutable
96Variables</b> - This chapter talks about adding user-defined local variables
Chris Lattner4134c282007-11-06 06:32:18 +000097along with an assignment operator. The interesting part about this is how
98easy and trivial it is to construct SSA form in LLVM: no, LLVM does <em>not</em>
99require your front-end to construct SSA form!</li>
Chris Lattner95ce0d62007-11-06 05:02:48 +0000100<li><b><a href="LangImpl8.html">Chapter #8</a>: Conclusion and other useful LLVM
101tidbits</b> - This chapter wraps up the series by talking about potential
102ways to extend the language, but also includes a bunch of pointers to info about
103"special topics" like adding garbage collection support, exceptions, debugging,
104support for "spaghetti stacks", and a bunch of other tips and tricks.</li>
105
106</ul>
107
Chris Lattner4134c282007-11-06 06:32:18 +0000108<p>By the end of the tutorial, we'll have written a bit less than 700 lines of
109non-comment, non-blank, lines of code. With this small amount of code, we'll
Chris Lattner95ce0d62007-11-06 05:02:48 +0000110have built up a very reasonable compiler for a non-trivial language including
111a hand-written lexer, parser, AST, as well as code generation support with a JIT
112compiler. While other systems may have interesting "hello world" tutorials,
113I think the breadth of this tutorial is a great testament to the strengths of
114LLVM and why you should consider it if you're interested in language or compiler
115design.</p>
116
117<p>A note about this tutorial: we expect you to extend the language and play
Chris Lattner4134c282007-11-06 06:32:18 +0000118with it on your own. Take the code and go crazy hacking away at it, compilers
119don't need to be scary creatures - it can be a lot of fun to play with
120languages!</p>
Chris Lattner95ce0d62007-11-06 05:02:48 +0000121
Chris Lattnerc38ef542007-10-22 04:32:37 +0000122</div>
123
124<!-- *********************************************************************** -->
Chris Lattner128eb862007-11-05 19:06:59 +0000125<div class="doc_section"><a name="language">The Basic Language</a></div>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000126<!-- *********************************************************************** -->
127
128<div class="doc_text">
129
130<p>This tutorial will be illustrated with a toy language that we'll call
Chris Lattner4134c282007-11-06 06:32:18 +0000131"<a href="http://en.wikipedia.org/wiki/Kaleidoscope">Kaleidoscope</a>" (derived
132from "meaning beautiful, form, and view").
Chris Lattnerc38ef542007-10-22 04:32:37 +0000133Kaleidoscope is a procedural language that allows you to define functions, use
134conditionals, math, etc. Over the course of the tutorial, we'll extend
Chris Lattner619bc0a2007-11-05 20:13:56 +0000135Kaleidoscope to support the if/then/else construct, a for loop, user defined
136operators, JIT compilation with a simple command line interface, etc.</p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000137
Chris Lattner619bc0a2007-11-05 20:13:56 +0000138<p>Because we want to keep things simple, the only datatype in Kaleidoscope is a
Chris Lattnerc38ef542007-10-22 04:32:37 +000013964-bit floating point type (aka 'double' in C parlance). As such, all values
140are implicitly double precision and the language doesn't require type
141declarations. This gives the language a very nice and simple syntax. For
Chris Lattner619bc0a2007-11-05 20:13:56 +0000142example, the following simple example computes <a
143href="http://en.wikipedia.org/wiki/Fibonacci_number">Fibonacci numbers:</a></p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000144
145<div class="doc_code">
146<pre>
147# Compute the x'th fibonacci number.
148def fib(x)
Chris Lattnere6c91042007-10-22 06:34:15 +0000149 if x &lt; 3 then
Chris Lattnerc38ef542007-10-22 04:32:37 +0000150 1
151 else
152 fib(x-1)+fib(x-2)
153
154# This expression will compute the 40th number.
155fib(40)
156</pre>
157</div>
158
Duncan Sands72261ff2007-11-05 16:04:58 +0000159<p>We also allow Kaleidoscope to call into standard library functions (the LLVM
Chris Lattnerc38ef542007-10-22 04:32:37 +0000160JIT makes this completely trivial). This means that you can use the 'extern'
161keyword to define a function before you use it (this is also useful for mutually
162recursive functions). For example:</p>
163
164<div class="doc_code">
165<pre>
166extern sin(arg);
167extern cos(arg);
168extern atan2(arg1 arg2);
169
170atan2(sin(.4), cos(42))
171</pre>
172</div>
173
Chris Lattner4134c282007-11-06 06:32:18 +0000174<p>A more interesting example is included in Chapter 6 where we write a little
175Kaleidoscope application that <a href="LangImpl6.html#example">displays
176a Mandelbrot Set</a> at various levels of magnification.</p>
177
178<p>Lets dive into the implementation of this language!</p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000179
180</div>
181
182<!-- *********************************************************************** -->
Chris Lattner128eb862007-11-05 19:06:59 +0000183<div class="doc_section"><a name="lexer">The Lexer</a></div>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000184<!-- *********************************************************************** -->
185
186<div class="doc_text">
187
188<p>When it comes to implementing a language, the first thing needed is
189the ability to process a text file and recognize what it says. The traditional
190way to do this is to use a "<a
191href="http://en.wikipedia.org/wiki/Lexical_analysis">lexer</a>" (aka 'scanner')
192to break the input up into "tokens". Each token returned by the lexer includes
193a token code and potentially some metadata (e.g. the numeric value of a number).
194First, we define the possibilities:
195</p>
196
197<div class="doc_code">
198<pre>
199// The lexer returns tokens [0-255] if it is an unknown character, otherwise one
200// of these for known things.
201enum Token {
202 tok_eof = -1,
203
204 // commands
205 tok_def = -2, tok_extern = -3,
206
207 // primary
208 tok_identifier = -4, tok_number = -5,
209};
210
211static std::string IdentifierStr; // Filled in if tok_identifier
212static double NumVal; // Filled in if tok_number
213</pre>
214</div>
215
216<p>Each token returned by our lexer will either be one of the Token enum values
Chris Lattner4134c282007-11-06 06:32:18 +0000217or it will be an 'unknown' character like '+', which is returned as its ASCII
Chris Lattnerc38ef542007-10-22 04:32:37 +0000218value. If the current token is an identifier, the <tt>IdentifierStr</tt>
219global variable holds the name of the identifier. If the current token is a
220numeric literal (like 1.0), <tt>NumVal</tt> holds its value. Note that we use
221global variables for simplicity, this is not the best choice for a real language
222implementation :).
223</p>
224
Chris Lattner619bc0a2007-11-05 20:13:56 +0000225<p>The actual implementation of the lexer is a single function named
226<tt>gettok</tt>. The <tt>gettok</tt> function is called to return the next token
227from standard input. Its definition starts as:</p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000228
229<div class="doc_code">
230<pre>
231/// gettok - Return the next token from standard input.
232static int gettok() {
233 static int LastChar = ' ';
234
235 // Skip any whitespace.
236 while (isspace(LastChar))
237 LastChar = getchar();
238</pre>
239</div>
240
241<p>
242<tt>gettok</tt> works by calling the C <tt>getchar()</tt> function to read
243characters one at a time from standard input. It eats them as it recognizes
Chris Lattner619bc0a2007-11-05 20:13:56 +0000244them and stores the last character read, but not processed, in LastChar. The
Chris Lattnerc38ef542007-10-22 04:32:37 +0000245first thing that it has to do is ignore whitespace between tokens. This is
246accomplished with the loop above.</p>
247
Chris Lattner619bc0a2007-11-05 20:13:56 +0000248<p>The next thing <tt>gettok</tt> needs to do is recognize identifiers and
249specific keywords like "def". Kaleidoscope does this with this simple loop:</p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000250
251<div class="doc_code">
252<pre>
253 if (isalpha(LastChar)) { // identifier: [a-zA-Z][a-zA-Z0-9]*
254 IdentifierStr = LastChar;
255 while (isalnum((LastChar = getchar())))
256 IdentifierStr += LastChar;
257
258 if (IdentifierStr == "def") return tok_def;
259 if (IdentifierStr == "extern") return tok_extern;
260 return tok_identifier;
261 }
262</pre>
263</div>
264
Chris Lattner619bc0a2007-11-05 20:13:56 +0000265<p>Note that this code sets the '<tt>IdentifierStr</tt>' global whenever it
266lexes an identifier. Also, since language keywords are matched by the same
267loop, we handle them here inline. Numeric values are similar:</p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000268
269<div class="doc_code">
270<pre>
271 if (isdigit(LastChar) || LastChar == '.') { // Number: [0-9.]+
272 std::string NumStr;
273 do {
274 NumStr += LastChar;
275 LastChar = getchar();
276 } while (isdigit(LastChar) || LastChar == '.');
277
278 NumVal = strtod(NumStr.c_str(), 0);
279 return tok_number;
280 }
281</pre>
282</div>
283
284<p>This is all pretty straight-forward code for processing input. When reading
285a numeric value from input, we use the C <tt>strtod</tt> function to convert it
286to a numeric value that we store in <tt>NumVal</tt>. Note that this isn't doing
Duncan Sands72261ff2007-11-05 16:04:58 +0000287sufficient error checking: it will incorrectly read "1.23.45.67" and handle it as
Chris Lattnerc38ef542007-10-22 04:32:37 +0000288if you typed in "1.23". Feel free to extend it :). Next we handle comments:
289</p>
290
291<div class="doc_code">
292<pre>
293 if (LastChar == '#') {
294 // Comment until end of line.
295 do LastChar = getchar();
296 while (LastChar != EOF &amp;&amp; LastChar != '\n' &amp; LastChar != '\r');
297
298 if (LastChar != EOF)
299 return gettok();
300 }
301</pre>
302</div>
303
Chris Lattner71155212007-11-06 01:39:12 +0000304<p>We handle comments by skipping to the end of the line and then return the
Chris Lattner4134c282007-11-06 06:32:18 +0000305next token. Finally, if the input doesn't match one of the above cases, it is
306either an operator character like '+' or the end of the file. These are handled
307with this code:</p>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000308
309<div class="doc_code">
310<pre>
311 // Check for end of file. Don't eat the EOF.
312 if (LastChar == EOF)
313 return tok_eof;
314
315 // Otherwise, just return the character as its ascii value.
316 int ThisChar = LastChar;
317 LastChar = getchar();
318 return ThisChar;
319}
320</pre>
321</div>
322
Chris Lattner619bc0a2007-11-05 20:13:56 +0000323<p>With this, we have the complete lexer for the basic Kaleidoscope language
324(the <a href="LangImpl2.html#code">full code listing</a> for the Lexer is
325available in the <a href="LangImpl2.html">next chapter</a> of the tutorial).
Chris Lattnerc38ef542007-10-22 04:32:37 +0000326Next we'll <a href="LangImpl2.html">build a simple parser that uses this to
Chris Lattnere6c91042007-10-22 06:34:15 +0000327build an Abstract Syntax Tree</a>. When we have that, we'll include a driver
328so that you can use the lexer and parser together.
Chris Lattnerc38ef542007-10-22 04:32:37 +0000329</p>
330
331</div>
332
333<!-- *********************************************************************** -->
334<hr>
335<address>
336 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
337 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
338 <a href="http://validator.w3.org/check/referer"><img
Chris Lattner8eef4b22007-10-23 06:30:50 +0000339 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Chris Lattnerc38ef542007-10-22 04:32:37 +0000340
341 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
342 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
343 Last modified: $Date: 2007-10-17 11:05:13 -0700 (Wed, 17 Oct 2007) $
344</address>
345</body>
346</html>